Benzodiazepines are among the most widely abused drugs, and like other drugs of abuse, they hijack the brain's dopaminergic reward system, in which several GABAA receptor subtypes are expressed in different neuronal cell types. The modulation of the reward system and of benzodiazepine self-administration by distinct GABAA receptor subtypes is only poorly understood. Here, we want to test the hypothesis that GABAA receptors bidirectionally modulate these behaviors. In the absence of chemical compounds which are truly specific for a GABAA receptor subtype, we propose to use a novel combined genetic and pharmacological approach to create a model system which will enable highly specific modulation of the activity of individual GABAA receptor subtypes. This will be achieved by using the non- selective benzodiazepine drugs diazepam and midazolam in triple point-mutated mice, in which these drugs are a true ?1-specific, ? 2-specific, ? 3-specific, or ? 5-specific full agonists, respectively. This system will make it possible to test whether potentiatio of a particular GABAA receptor subtype is sufficient for reward enhancement and benzodiazepine self-administration. PUBLIC HEALTH RELEVANCE: The neurobiological mechanisms underlying benzodiazepine abuse are not well understood, and in particular, the roles of the individual GABAA receptor subtypes in reward-related and self-administration behaviors are largely unknown. We propose to examine whether any of the major four benzodiazepine- sensitive GABAA receptor subtypes is sufficient for modulation of these behavioral responses and plan to demonstrate that GABAA receptors bidirectionally modulate these processes. The knowledge of how each of these GABAA receptor subtypes affects the abuse potential of benzodiazepines is relevant for the development of GABAA receptor subtype-selective compounds for the treatment of anxiety, chronic pain, cognitive deficits, and of benzodiazepine abuse.